wednesday-git
Unified Git workflow. Manages the entire task lifecycle: branch creation (sprint), atomic commits (git-os), and PR opening (pr-create).
git clone --depth 1 https://github.com/wednesday-solutions/ai-agent-skills /tmp/wednesday-git && cp -r /tmp/wednesday-git/skills/wednesday-git ~/.claude/skills/wednesday-gitSKILL.md
# Wednesday Git Workflow (GIT-OS)
This skill manages the complete lifecycle of a feature or fix. It ensures that every branch, commit, and PR follows the project's quality and automation standards.
## Lifecycle Phases
### 1. Starting a Task (`sprint`)
When a dev provides a ticket and wants to start working:
- **Action**: Create a branch following the `type/name` format (e.g., `feat/user-auth`).
- **Rule**: Max 40 characters, kebab-case, no ticket numbers in the branch name.
- **Tool**: `Bash(git checkout -b <branch_name>)`
### 2. Committing Changes (`git-os`)
When code is ready for an atomic commit:
- **Format**: `type(scope?): Description` (e.g., `feat(auth): Add login endpoint`).
- **Types**: `feat`, `fix`, `refactor`, `perf`, `docs`, `style`, `test`, `chore`.
- **Atomic Rule**: One logical change per commit. If you've done multiple things, split them into separate commits.
- **Human-Authored**: NEVER include AI attribution or fingerprints.
- **Linting**: Before committing, ensure `npx --no-install commitlint` would pass.
### 3. Opening a Pull Request (`pr-create`)
When the task is finished and ready for review:
- **Pre-Push Checklist**: Run `npm run lint`, `format:check`, `test`, and `build`.
- **Validation**:
- **Max 6 files per PR**: If the PR is larger, suggest splitting it.
- **Conventional Title**: Title must match `type(scope): Description`.
- **Metadata**: Extract ticket ID from branch name (e.g., `WED-142`).
- **PR Description Template**:
```markdown
### Ticket Link
---
### Description
---
### Steps to Test
---
### GIFs (if applicable)
---
```
- **Push & Create**:
1. `Bash(git push origin <branch>)`
2. `Bash(gh pr create --title "<title>" --body "<body>")` using the template above.
---
## 🚫 Never
- **Direct Commits**: Never commit directly to `main` or `develop`.
- **Dirty History**: Never bundle multiple concerns into one commit.
- **Skip Checks**: Never open a PR without running the full pre-push checklist.
- **AI Fingerprints**: Never use "AI-generated" descriptions or co-author tags.
## ⚠️ Safety Thresholds
- **Risk > 80**: If the `codebase-intel` risk score is > 80, require explicit confirmation before pushing.
- **Quality Gate**: If Sonar is enabled and the quality gate fails, STOP — do not push.
## 🛠 Tools
- `Bash` for all `git` and `gh` operations.
- `Read` for `package.json` and `.wednesday/config.json` to check for project-specific automation.Use when asked to audit a module, check its health, or assess whether it should be refactored. Runs structural query, risk check, and test generation automatically.
Use when a dev is new to the project, asks for an overview, or wants to get oriented quickly. Fills graph gaps then runs a guided codebase interview.
Full PR review orchestrator. Use when asked to review a PR or check if it is ready to merge. Runs blast radius + drift check on changed files, then hands off to pr-review for comment triage and fixes. Do NOT use pr-review directly for full reviews — use this.
Enforces architecture boundaries defined in PLAN.md. Use when a PR crosses module/service boundaries, when the dev asks "are we following the architecture?", or as a scheduled architecture health check. Not for querying what a module does — use brownfield-chat for that.
Unified codebase intelligence. Handles all questions about structure, logic, risk, and dependencies. Combines natural-language Q&A with deterministic lookups and pre-edit blast radius checks.
Pre-deploy and post-deploy checklist skill. Ensures env vars, migrations, CI, rollback plan, smoke tests, and monitoring are verified before and after every deployment.
Parallel persona planning for new projects. Research agent runs first to build domain context, then Architect, PM, and Security agents run in parallel. Synthesis agent combines all perspectives into a detailed GSD-style PLAN.md with Tensions section.
Fix engine for PR review comments. Fetches review comments (Gemini bot or human), categorizes by impact, posts a prioritized fix queue, and applies fixes on dev approval. Called directly for quick fixes, or internally by pr-review-agent as part of full PR review.